Aye, probably better to have a ui in box to enable/disable and whitelist domains. I don't have a corporate customer waiting for that, I just thought I left the suggestion here for the future, maybe it could be added when something else in that department is touched. 😉
@robi or even Jirafeau which is available on Cloudron and does the job nicely.
I use that, but also have greatly increased the maximum too because sometimes I e.g. want to email a large ebook to Amazon to automatically get it on my Kindle (although sometimes it's too large and Amazon reject it at their end - need to check what their max is).
If this is disabled or archived, this also means that it is still "known" to the system, currently the data simply is not removed from the disk, but other relevant information is purged from the database.
So I agree, that deleting an account should also delete the mailbox data, if there is no proper way to restore that (currently you have to create a new mailbox record with the same mailbox handle and magically the data comes back). For that some archived state could be the solution to have a structured way to restore.
Correct. We don't need to re-create container to change the memory limit. But we still need to restart it after adjusting memory limit because of the limitations in our packaging.
I think over time we have learnt that it is not a good idea to setup apache to auto-scale based on container memory limit. Those things are very dependent on the app/plugin use. I think java apps require a restart at this point since the JVM gets passed in the heap memory as a flag on startup, maybe there is a workaround for this as well, have to investigate.
But at that point, we can atleast make the memory limit code not re-create container which I think is where bulk of the slowness is.
@nebulon For sure, yeah I figured it may need more interest before it's implemented, and realistically I'm probably the only Cloudron user who's hosting on LunaNode (though I'd love to know if others are too), as I think LunaNode is a nice hidden gem that's not overly used yet and I switched to them recently as I had heard great things and their pricing is pretty good. Brief recommendation at the bottom for anyone interested. I'll let them know too about Cloudron in case they can do anything to help out too. 👍
On the topic of DNS at LunaNode though... it seems pretty awesome from what I can tell, as it has integrated load balancing and failover which is perfect and works with their included monitoring feature too (which is like Uptime Robot but provided by LunaNode), and so if my monitor fails and I have a second DNS entry, it'll automatically remove the failed one from the DNS records temporarily until it's back online, giving only the backup DNS entry for the "hot standby" server. Hoping to take advantage of those features in the not too distant future if it's feasible for me to run a second "hot standby" server. Waiting on Cloudron to add in some functionality to clustering which I think is coming in v6. 😉 Once that's there, I hope to be able to do that. Nudge nudge. haha.
🗣 Brief recommendation based on my initial tests so far for other Cloudron users who want a good inexpensive hosting provider: I recently moved my Cloudron server from OVH to LunaNode to test out their performance, and the exact same Cloudron server in OVH used to take 10 minutes to fully come online after a reboot in OVH, and having restored that same server in LunaNode it now takes only 2-3 minutes to fully come online. That's also with some memory-intensive applications like Mastodon and NodeBB (which are always the last two to start in my environment given their resource usage). Same 8 GB nodes between OVH and LunaNode too. This performance improvement now cuts my downtime after reboots for security updates by almost 75%! That's my brief recommendation for LunaNode. haha.
After thinking about it, the graph chosen is better to view memory usage (which fluctuate much more than disk).
I would then suggest separating both information:
keep the current graph for the app usage (the Y axis adapting to the app using the most memory)
add a bar for total memory used, modelled on the total disk usage (could be a single colour for memory usage)
total disk usage
This would help maximize the amount of information you can visualize in one go and help detect spikes better.
@murgero (you probably know this but just writing it out for the wider audience)...
Cloudron was primarily made to install and manage pre-packaged apps. This meant that addons were defined by the package and we don't want end users to make difficult decisions as to what database to use and what caching backend to use. Lesser choices cover 90% of the use cases.
But over time, we saw a lot of demand about custom apps. And thus LAMP app and even allowing custom packages. And now we probably also need dynamic addons (ie. select which database post installation or just before installation). While this is doable, this moves Cloudron more and more into PaaS territory. My initial instinct is that it's fine to venture into this space and maybe cover 90% of the use cases but we have to think through it 🙂
@girish Thanks for the answer, but I didn't mean the internal server. I would like to install MySQL or another standalone server. I would like to try some things. Try and Error 🙂 I have three instances of Cloudron and I would like to install my own applications and if I can understand all things and requirements ... I do not know what can happen ....
First I have successfull ended a course on udemy.com for Docker.
This is possible but will increase complexity in deploys and maintenance. I have also noticed that MySQL changes things between patch releases a lot. Maybe because their release numbers don't follow semver .